What the AI?!...Equipped.ai
As the AI trend continues to transform the PE industry, we speak to Equipped.ai’s CEO Ed Green, who tells us more about how the firm is integrating AI into its product offerings.
The data analytics and technology platform originally spun out of AnaCap in 2021 and provides consultancy and software services.
Its software – Minerva – consolidates investment, portfolio management and fundraising processes into a single, cloud-based system that includes dashboards, workflow tools and automated reporting.
The Drawdown (TDD): What does AI mean for Equipped.ai?
Ed Green (EG): It seems that every single private equity fund is talking about AI adoption, and everyone has a sense of wanting to get to the front of the pack.
When we get into the detail of the readiness of those fund managers for AI, we find a spectrum of firms. We ask the following questions: Is there a standardised data set in place? Is there data integrity? Is there a single source of truth? Then, we ask whether that firm can draw meaningful analytics from its own data. Once that robust dataset has been established, we can then open up new opportunities through analytics, intelligent automation and machine learning, where at each stage the level of automation and autonomy of the technology is increasing. The last stage of the process is understanding how AI can be used to unlock additional insights from large language models.
We often see firms that are really at the starting point of that journey. They've raised capital, deploying it through an effective strategy, and are now looking to unlock the efficiencies from their data assets.
We use analytical intelligence to help firms at the very first stages of that spectrum. As the name suggests, we centre ourselves around equipping fund managers for that technological revolution created by high performance computing and AI.
TDD: How do you use AI to help streamline operations for PE firms?
EG: GPs tend to get separate information from fund administrators, accounting systems and bank accounts, trying to make sense of it all.
Equipped.ai uses its own algorithms to piece the data together. We apply AI and our bespoke algorithms to help match and synchronise all those disparate datasets. For example, we have one client with hundreds of thousands of credit positions in Spain with a range of external servicers who provide them with regular reporting. A separate set of reporting comes through their own banking records. We help them automate the process of comparing those statements with detailed reporting from the multiple local providers in Spain.
TDD: And do you have any examples of Equipped.AI helping firms that are in the later stages of AI adoption?
EG: Yes, we're currently piloting a 2.0 tool within our Minerva system; the tool is akin to ChatGPT. It aims to automate the process of writing small reports to summarise investment performance to partners and the investment committee.
The idea is that the report is written using a tightly managed framework, flagging up all the issues and exceptions for the end user. The tool should also help investment managers navigate the Minerva system.
TDD: What risk factors have you encountered when piloting the tool?
EG: I became aware of certain generative AI risk factors when I used ChatGPT to help me with a biography I was writing. The biography was roughly 40% accurate and 60% totally made up. It was fascinating to unearth a really interesting overview of a person that resembled me but with a very different career path and educational background!
This is a great example of how the process and the governance around how you use these tools needs to be very thorough.
You can easily avoid these so-called ‘hallucination’ risks by asking the tool very specific questions. You also need to create secure and structured interfaces, so that the tool only has access to very specific information, rather than exposing it to broader subsets.
The right parameterisation and structuring of the questions you ask can help prevent hallucinations from occurring.
Another risk is that the AI delivers a set of results that appear to be empirical but incorporate unintentional biases. So, we're setting up frameworks for testing those outputs and we are ensuring that there is a programme of regular testing and audits in place as we go forward.
Finally, I think transparency and explainability is very important for any AI tool adoption. Within Minerva, we have the capability to show a particular data field and display the component parts of the calculation that have gone into that because every single CFO simply wants to have a clear audit trail.
TDD: Should we be excited or worried about AI developments?
EG: In terms of its application in the alternative investment industry, I think we should definitely be excited. In the work we do for clients, AI has the ability to replace dull, repetitive, manual tasks and free up resources for valuable ones that reward talent and innovation. If the finance team or the operations team are able to free themselves up from doing all that nitty-gritty work, then they can impact bigger ticket items and deliver more material value- add tasks moving forward.
To read more interviews in the What the AI?!... series, click here